|
Sandra (the System
ANalyzer, Diagnostic and Reporting Assistant)
is an information and diagnostic utility put out by the
folks at SiSoftware. Besides benchmarking, it provides a
host of other information about your hardware and
software.
 |
|
SiSoftware
Sandra |
|
Our Sandra Doing
Her Thing! |
|
We began our
testing with four of the built-in sub-system tests that
are part of SiSoftware's Sandra 2002 benchmarking suite
(CPU, Multimedia, Memory and File System), running at the
CPU's default clockspeed of 1666MHz. (12.5x133) and an
overclocked speed of 1750MHz. (12.5x140).
CPU @ 1666MHz.
CPU @ 1750MHz.
If you're used
to seeing Sandra CPU scores from single CPU machines,
these numbers will surely make your jaw drop! Even
at default clockspeeds, ALU performance dominated every
other reference system listed. The dual 2GHz. Xeon
machine took the FPU tests though. When we
overclocked the the K7D Master, we saw a significant jump
in performance. The Dual Athlon machine continued to
pull away in the ALU test, while their FPU performance
came within a hair of the Xeon machine.
MM @ 1666MHz.
MM @ 1750MHz.
The Sandra
Multimedia tests belong the Athlon no matter which way you
slice it. At both the default and overclocked
speeds, the Dual Athlons bested every other reference
system listed. In fact, we had to select an 8-Way
Xeon rig from the reference list before anything in
Sandra's database could beat our dual overclocked 2000+
CPUs!
Hard Drive
As you can see
in the graph above, hard drive performance with AMD's
Southbridge left a little to be desired. The same
IBM hard drive we used on this consistently scores
on the 20K+ range on other test machines. 17K+ is s
decent score, but it is approximately 20% lower than what
we expected from this drive. Perhaps with a little
driver tuning AMD can remedy this situation in the future.
Memory @
1666MHz. Memory @
1750MHz.
The all (well,
almost all) important memory bandwidth test shows AMD's
762 chipset performing nicely, but not quite on par with
even the older AMD 760 chipset. Even while
overclocked to a 140MHz. FSB, with the memory set at the
most aggressive timings available, the MSI K7D Master
barely managed to catch the AMD 760 reference machine.
Systems using nForce or KT333 chipsets would have an even
larger lead in this department.
|
|
|
MadOnion PC Mark 2002 |
|
Another
Youngster... |
|
MadOnion's new
PCMark 2002 benchmarking suite, is rapidly gaining
acceptance amongst on line PC Hardware Test Publications.
PCMark 2002 is very simple to run, and produces repeatable
results. We ran their "CPU" and "Memory" performance
modules, which incorporate the following tests:
CPU Test:
Memory Test
Technical details: (Quoted)
Raw read,
write, and read-modify-write operations are performed
starting from a 3072 kilobytes array decreasing in size to
1536 KB, 384 KB, 48 KB and finally 6 KB. Each size of
block is tested two second and the amount of accessed data
is given as result. In the STL container test a list of
116 byte elements is constructed and sorted by an integer
pseudo-random key. The list is then iterated through as
many times as possible for 2 seconds and the total size of
the accessed elements is given as result. There are 6 runs
of this test, with 24576 items in the largest run
corresponding to a total data amount of 1536 KB,
decreasing in size to 12288 items (768 KB), 6144 items
(384 KB), 1536 items (96 KB), 768 items (48 KB) and 96
items in the smallest run corresponding to 6 KB of total
data.
The CPU score
is in line with a single Athlon XP 2000+ processor.
PCMark 2002 is obviously not multithreaded. If it
was, the CPU score would have been much higher. The
Memory score may be a little lower than you're accustomed
to seeing if you've read recent reviews of VIA KT333 based
systems. Our Abit AT7 system that we'll be using for
reference numbers from this point forward scored over 3400
in this same test.
For the
remainder is this review, we'll be comparing the dual
Athlon MP 2000+ equipped MSI K7D Master to the Abit AT7,
Abit's "Legacy Free" motherboard based on VIA's KT333
chipset using an Athlon XP 2100+.
|
|
|
CliBench
MK III SMP 0.7.15 |
|
New Edition |
|
Finding
benchmarks that accurately demonstrate the benefits of an
SMP system is no easy task. Thankfully the gurus at
2CPU
have found a few gems, one of which is CliBench.
CliBench is a benchmark suite, portions of which were self
developed by Daemonware, while others are slightly
modified versions of industry standard benchmarks. The
"simple" CPU intensive tests clearly show the differences
between architectures.
CliBench is a small download (it will actually fit on
a floppy), try it out for yourself and see how your
personal rig compares to ours. Users can check their
CPU, FPU, memory and hard drive performance, and it is
fully multithreaded, so you can easily check SMP systems
with up to 128 processors!
The CliBench
scores clearly show the Dual-CPU rig almost doubling the
performance of the single CPU system in every test, with
the exception Memory throughput. When used in
conjunction with quality PC2700 DDR RAM, the VIA KT333
really shines in the memory bandwidth department.
|
|
|
Flask MPEG Encoding (DivX) |
|
SMP Capable |
|
Flask MPEG is
a Digital Video Stream conversion utility that takes a
video of a certain file type, and converts it to another
format. We took a 19MB MPEG 2 video clip (that
made it's debut in Davo's 533Mhz. BUS P4 review) and
converted it to DIVX format using the DIVX Pro 5.01 Codec.
Here are the results.
Looking at the
graph, it's obvious that Flask MPEG is capable of using
the power of the second CPU installed in the MSI K7D
Master. Having the extra horsepower of a second CPU
in the system resulted in roughly a 63% performance
advantage for the Dual-CPU system. In fact, at 33.56
FPS, the K7D Master with dual Athlon MP 2000+ CPUs
outperformed the fastest single processor system we have
ever tested here on HotHardware, a 2.53GHz. P4 with PC1066
RDRAM, by 31%,
The Stones,
Quake 3 and the Rating
| 
